Implementation of a finite volume method and calculation of flows of a viscous compressible gas on graphics processor units

Authors

Keywords:

graphics processor unit, parallel algorithm, gas dynamics, finite volume method, unstructured mesh, CUDA technology

Abstract

Possibilities of the use of general purpose graphics processor units for the numerical simulation of flows of a viscous compressible gas are discussed. A finite volume method is applied to solve the Navier-Stokes equations on unstructured meshes. The CUDA technology is used to implement parallel computational algorithms. The solution of some gas dynamics problems on graphics processor units is presented. A number of approaches to the optimization of program codes related to the use of different types of memory are discussed. The speed-up of solving procedures on graphics processor units is compared to their implementations on central processor units with the use of various meshes and various methods of distribution of input data into blocks.

Author Biographies

K.N. Volkov

V.N. Emelyanov

P.G. Smirnov

I.V. Teterina

A.G. Karpenko

St Petersburg University,
Mathematics and Mechanics Faculty
Universitetsky prospekt, 28, Peterhof, St. Petersburg, 198504, Russia
• PhD Student

References

  1. Волков К.Н., Емельянов В.Н., Карпенко А.Г., Курова И.В., Серов А.Е., Смирнов П.Г. Численное решение задач гидродинамики на графических процессорах общего назначения // Вычислительные методы и программирование. 2013. 14, № 1. 82-90.
  2. Brandvik T., Pullan G. Acceleration of a 3D Euler solver using commodity graphics hardware // AIAA Paper. 2008. N 2008-607.
  3. Corrigan A., Camelli F., LHohner R., Wallin J. Running unstructured grid-based CFD solvers on modern graphics hardware // AIAA Paper. 2009. N 2009-4001.
  4. Corrigan A., Camelli F., LHohner R., Mut F. Semi-automatic porting of a large-scale Fortran CFD code to GPUs // Int. J. for Numerical Methods in Fluids. 2012. 69, N 2. 314-331.
  5. Brandvik T., Pullan G. An accelerated 3D Navier-Stokes solver for flows in turbomachines // ASME Paper. 2009. N GT2009-60052.
  6. Kampolis I.C., Trompoukis X.S., Asouti V.G., Giannakoglou K.C. CFD-based analysis and two-level aerodynamic optimization on graphics processing units // Computer Methods in Applied Mechanics and Engineering. 2010. 199, N 9-12. 712-722.
  7. Appleyard J., Drikakis D. Higher-order CFD and interface tracking methods on highly-parallel MPI and GPU systems // Computers and Fluids. 2011. 46, N 1. 101-105.
  8. Волков К.Н., Емельянов В.Н. Моделирование крупных вихрей в расчетах турбулентных течений. М.: Физматлит, 2008.
  9. Волков К.Н. Дискретизация уравнений Навье-Стокса на неструктурированной сетке при помощи метода контрольного объема и разностных схем повышенной разрешающей способности // Ж. вычислит. матем. и матем. физики. 2008. 48, № 7. 1250-1273.
  10. Roe P.L. Characteristic-based schemes for the Euler equations // Annual Review of Fluid Mechanics. 1986. 18. 337-365.
  11. Harten A. High resolution schemes for hyperbolic conservation laws // J. of Computational Physics. 1983. 49, N 3. 357-393.
  12. Sod G.A. A survey of several finite difference methods of systems of nonlinear hyperbolic conservation laws // J. of Computational Physics. 1978. 27, N 1. 1-31.

Published

10-04-2013

How to Cite

Волков К., Емельянов В., Смирнов П., Тетерина И., Карпенко А. Implementation of a Finite Volume Method and Calculation of Flows of a Viscous Compressible Gas on Graphics Processor Units // Numerical Methods and Programming (Vychislitel’nye Metody i Programmirovanie). 2013. 14. 183-194

Issue

Section

Section 1. Numerical methods and applications

Most read articles by the same author(s)

1 2 3 4 > >>